What are the significant roles of siderophores?

Siderophores are low molecular weight, high affinity iron chelating molecules that are essential virulence factors in many Gram-negative bacterial pathogens. Whereas the chemical structure of siderophores is extremely variable, the function of siderophores has been narrowly defined as the chelation and delivery of iron to bacteria for proliferation. The discovery of the host protein Lipocalin 2, capable of specifically sequestering the siderophore Enterobactin but not its glycosylated-derivative Salmochelin, indicated that diversity in structure could be an immune evasion mechanism that provides functional redundancy during infection. However, there is growing evidence that siderophores are specialized in their iron-acquisition functions, can perturb iron homeostasis in their hosts, and even bind non-iron metals to promote bacterial fitness. The combination of siderophores produced by a pathogen can enable inter-bacterial competition, modulate host cellular pathways, and determine the bacterial “replicative niche” during infection. This review will examine both classical and novel functions of siderophores to address the concept that siderophores are non-redundant virulence factors used to enhance bacterial pathogenesis.

Specialization of siderophore function during infection: Salmochelin (Sal) evades Lipocalin 2 (Lcn2) to scavenge iron, Enterobactin (Ent) promotes lung invasion, and Yersiniabactin (Ybt) binds copper for detoxification and zinc for nutrition.

1. Introduction

Iron is required for many processes necessary for life, including DNA replication and electron transfer, due to its ability to assume multiple oxidative states.1–3 Each oxidation state poses challenges to the cell: ferrous iron (Fe2+) is highly toxic in its free form due to its participation in the Fenton reaction, whereas ferric iron (Fe3+) is insoluble at physiological pH and not readily bioavailable.4–6 Therefore, free iron levels are exceedingly low, and the majority of iron in the human body is bound by storage, transport, and metabolic proteins. Because bacteria require iron for replication during colonization and infection, pathogens must compete to acquire iron that is tightly regulated by the host. In order to outcompete tight iron binding by host molecules such as transferrin and lactoferrin, Gram-negative bacteria, Gram-positive bacteria, and some fungi secrete small iron-sequestering molecules called siderophores.4,5,7

Siderophores were first discovered in the 1950s with the identification of mycobactin as a growth factor for Mycobacterium johnei.8,9 Since then, over 500 distinct siderophores have been identified,10 indicating the importance of siderophores and the evolutionary pressure on pathogens to acquire iron from host environments. Siderophores are small molecules, often between 500–1500 daltons in molecular weight with high affinities for ferric iron.6,10 Many siderophores are important virulence factors, particularly in pathogens that encode multiple siderophores due to acquisition of siderophore synthesis systems by horizontal gene transfer.11,12 In fact, strains capable of over-producing siderophores are considered to be hypervirulent,13 whereas strains unable to produce or secrete siderophores have decreased virulence and fitness during infection and colonization.14–17

Though it is widely accepted that iron chelation by siderophores represents an important fitness advantage and virulence factor for many bacteria, the fact that so many unique siderophores are produced to perform one redundant function warrants further examination. If all siderophores chelate iron, what are the advantages of secreting more than one? The discovery of Lipocalin 2 (Lcn2; also siderocalin, neutrophil gelatinase-associated Lipocalin [NGAL], or 24p3), a molecule capable of specifically sequestering the prototypic catecholate siderophore Enterobactin (Ent),18 supported the idea that structural diversity with functional redundancy may allow evasion of host immune factors. Specifically, many enteric pathogens can produce a glycosylated-derivate of Ent called Salmochelin (Sal) that Lcn2 cannot bind.19 However, there is growing evidence that structural diversity allows specialization of function in scavenging iron and even other metals. These specialized roles of siderophores may impact virulence and alter the pathogenesis of bacterial infections.

This review will focus on the functional specialization of siderophores of Gram-negative pathogens and the impact of specialization on pathogenesis. First, the diverse capabilities of iron acquisition by siderophores, based on iron affinity, membrane permeability, and interactions with Lcn2 will be discussed. Next, siderophore functions other than iron scavenging, including affinity for other metals, heavy metal detoxification, and the ability of siderophores to perturb host pathways will be explored. Finally, we will examine the effect of the diverse siderophore structure and function on bacterial competition, pathogenesis, and virulence. These examples indicate that siderophores are distinct and non-redundant virulence factors, consistent with their structural diversity and production in specific combinations by successful pathogens.

2. Specialization of siderophores for iron delivery

In order to successfully scavenge iron and deliver it to its bacteria of origin, siderophores must outcompete host proteins and the iron-acquisition mechanisms of other microorganisms occupying the same niche. Variation in siderophore structure leads to differences in iron affinity, pH optimum, membrane partitioning, and ability to evade Lcn2 within the host.

2.1. Structural families and ferric iron (Fe3+) affinity

Bacterial siderophores can be divided into three major families based on the chemical groups involved in iron binding: catecholate, hydroxamate, and carboxylate.6,7 All families utilize negatively charged oxygen atoms to coordinate ferric iron, but each family has distinct characteristics that affect their affinity for iron. In addition to these three families, many mixed-type siderophores have been characterized, complicating the simple “three family” view of siderophores. An overview of the siderophore structural families is depicted in Fig. 1.

Siderophores can be divided into structural families. Siderophores can be divided into three main structural families: catecholates, hydroxamates, and carboxylates. Binding moieties for prototypic siderophores are highlighted in green (catecholate), blue (hydroxamate), and red (carboxylate). “Mixed type” siderophores are siderophores that are considered to have elements of two or more siderophore families. Binding moieties for “mixed type” siderophores are highlighted in orange (phenolate) or a color stated above, representing its binding moiety.6

One well-studied family of siderophores is the catecholate siderophores, which include Ent, its glycosylated-derivative Sal, and bacillibactin.20–22 Catecholate-based siderophores form 5-membered chelate rings and have the highest ferric iron affinity of any family.6,10 In particular, the Fe–Ent complex has a dissociation constant (Kd) of 10−49 M, and is capable of dissociating iron from host proteins, including the high affinity iron-binding protein transferrin (Kd = 10−20 M).6,23–25 The hydroxamate family of siderophores includes the fungal siderophores ferrichrome and (des)ferrioxamine (DFO; Kd = 10−30 M), a Streptomyces-derived siderophore used clinically to chelate iron.26–30 Hydroxamate siderophores also form five-membered chelate rings but have lower affinities for iron than catecholate siderophores.31

Carboxylate siderophores include staphyloferrin and citrate.32,33 In general, at physiological pH, carboxylate siderophores are less successful at iron chelation than both the catecholate and hydroxamate families. Instead, carboxylate siderophores are more efficient chelating ferric iron at acidic pH where catecholate and hydroxamate siderophores remain protonated.31 Thus, carboxylate siderophores may be useful to microbes living in acidic environments but are predicted to be outcompeted for iron scavenging by catechol-based siderophores in human serum and blood (pH 7.4).31 This is the case for phytosiderophores, which are plant-produced siderophores that utilize carboxylate oxygens to chelate iron in soil, where high affinity at pH 7 is less important than affinity at acidic pH.34

Some siderophores, such as yersiniabactin and pyochelin, contain phenol groups and are considered to be in a phenolate family of siderophores.35,36 However, rather than considering phenolate siderophores as a structural family, most consider these siderophores to be “mixed-type” siderophores, with elements of two or more siderophore families.6 Yersiniabactin (Ybt) has a moderate affinity for iron (Kd = 10−36.6 M)37 and includes phenolate, thiazole, oxazoline, and carboxylate groups involved in iron binding.6 Parabactin and carboxymycobactin are mixed type siderophores that include oxazoline rings.9 Other siderophores included in this group are aerobactin (Aer, Kd = 10−27 M)6,38 that contains hydroxamate and carboxylate moieties, petrobactin that contains catechol and carboxylate moieties and pyoverdin that is a complex mixed-type siderophore that expresses a yellow-green fluorescence.31,38–40

2.2. Lipocalin 2 binding and evasion

As the molecule with the highest known affinity for ferric iron, Ent represents a huge challenge to the host. To oppose the acquisition of iron by Ent, host epithelial cells and neutrophils secrete the protein Lcn2. Initially discovered as a neutrophil granule protein,41 it was later shown to bind to and sequester Ent with high specificity and affinity.14,18,42,43 The affinity of Lcn2 for Ent is similar to that of its bacterial receptor FepA, allowing Lcn2 to effectively compete with bacteria for binding to Ent, thus limiting bacterial growth.19,44 Lcn2 is induced as a general response to both Gram-positive and Gram-negative infections.42,45 Its protective effect was first illustrated in an Escherichia coli sepsis model, in which wild-type mice survived infection with an Ent-producing strain of E. coli but Lcn2-deficient mice did not. In contrast, Lcn2-deficient mice were equally as sensitive as wild-type mice to infection with Staphylococcus aureus, bacteria that do not rely on Ent for iron acquisition.42 Lcn2 is also protective against nasal colonization and pneumonia caused by Klebsiella pneumoniae and urinary tract infections caused by E. coli, provided that the bacteria depend on Ent for iron acquisition.12,14,46–48 In addition to Ent, Lcn2 can bind and sequester the catecholate siderophore bacillibactin from Bacillus anthracis and the mixed-type, oxazoline containing siderophores carboxymycobactin from M. tuberculosis and parabactin, demonstrating that Lcn2 is capable of binding siderophores from both Gram-positive and Gram-negative pathogens.49,50

Due to the strong bacteriostatic effects of Ent sequestration by Lcn2, bacteria have developed mechanisms to evade Lcn2 in their quest to obtain iron. One mechanism of evasion is to modify Ent through the addition of glucose groups to produce Sal.17,19,51 Sal is structurally identical to Ent except for the covalent addition of one to three glucose groups at the ends of the catecholate arms. The iroA gene cluster, found in many pathogenic Gram-negative bacteria, encodes the glycosylation, export, import, and esterase activities required to produce and utilize this siderophore.17,52 Glycosylation of Ent prevents interaction between Sal and Lcn2 due to steric hindrance, which leads to effective evasion of Lcn2, acquisition of iron required for bacterial replication, and the restoration of virulence in Lcn2-producing mice.14,17,19

Another method bacteria employ to evade Lcn2 is to encode additional siderophores that Lcn2 cannot bind. Strains of K. pneumoniae that encode Ybt readily cause pneumonia, whereas strains that only encode Ent are opportunistic, causing severe pneumonia but only in Lcn2-deficient mice.12 Similarly, Bacillus anthracis commonly secrete petrobactin in addition to bacillibactin, a Lcn2 ligand. Only petrobactin, a Lcn2-evasive siderophore, is required for virulence in macrophages and murine models.49,53–55 Many successful pathogenic bacteria encode multiple siderophores, including- or sometimes exclusively-Lcn2-evasive siderophores.12,56–59 For example, Yersinia pestis secretes Ybt but not Ent.60

Finally, an additional way to evade Lcn2 is to utilize xenosiderophores: siderophores produced by another microorganism that the pathogen cannot make itself. This has been demonstrated both in vitro and in vivo with bacterial use of the fungal siderophore ferrichrome.42 Addition of ferrichrome to E. coli H9049 enhances bacterial growth under iron-limiting conditions as well as in the presence of Lcn2. During infection models, wild-type mice infected with E. coli H9049 survived infection, but injection of ferrichrome caused susceptibility to infection.42 In sites of polymicrobial colonization or infection, the ability to utilize xenosiderophores could provide a general fitness advantage in competition with other microorganisms for iron and specifically aid in the evasion of Lcn2. Together, the literature demonstrates the significance of the siderophore–Lcn2 interaction, which causes detrimental effects to bacterial growth and infection but can be subverted by siderophore modification or the encoding of additional Lcn2-evasive siderophores.

2.3. Membrane partitioning

In the host, there are both intracellular and extracellular sources of iron, separated by cellular membranes. Differences in structure between various siderophores suggest disparities in siderophore function and ability, including the ability to access cellular contents via membrane partitioning.7 As reviewed above, the addition of hydrophilic glucosyl groups to Ent creates Sal, an Lcn2-evasive siderophore. Glycosylation, followed by linearization of Sal, also reduces the molecule's affinity for membranes, effectively decreasing the membrane partitioning constant.52 In contrast, Ent's high affinity for membranes causes it to partition into phospholipid vesicles. When iron is present outside these vesicles, the ability of Ent to scavenge this iron decreases ∼75%, whereas Sal retains its scavenging ability. Sal could therefore aid in the acquisition of extracellular iron in membrane-rich environments.17,52 Consistent with these differences in membrane partitioning, Ent depletes the labile iron pool of respiratory epithelial cells, whereas purified Sal has little effect.61

Although Ent readily partitions into membranes and depletes cellular iron, it does not appear to deliver cell-associated iron back to bacteria. In contrast, Aer secreted by E. coli can readily deliver host cellular iron back to bacteria.62 When E. coli were incubated with 59Fe labeled transferrin or leukemia cells, but separated from both by a dialysis membrane, Ent delivered iron by scavenging it away from transferrin whereas Aer scavenged iron primarily from host cells.62 The preferential scavenging of iron from tissues by Aer may explain why invasive strains of bacteria are more likely to secrete Aer even though it has lower affinity for iron than other siderophores.62,63 Although these data indicate that Aer efficiently accesses cellular iron, it is not known if this corresponds to a high membrane partitioning constant. The differing ability of siderophores to chelate iron among physiologically relevant molecules, such as lipids, indicates distinct roles for siderophores depending on the niche in which the bacterium finds itself.

3. Non-iron delivery functions of siderophores

As detailed above, siderophores are highly potent iron binding molecules that are specialized in their ability to scavenge iron for bacterial replication under varying conditions. Recent literature suggests novel roles and characteristics of siderophores, including the ability to bind other metals, prevent oxidative burst during infection, and perturb host cell homeostasis (Fig. 2).

Novel characteristics and roles of siderophores. (A) Bacteria can secrete Ybt to sequester Zn2+, which is taken up by the inner membrane receptor YbtX. This function can compensate for loss of the Zn2+ active transport system ZnuABC in septicemic plague.79 (B) Cu2+–Ybt can act as a superoxide dismutase mimic within phagosomes, protecting bacteria from harmful reactive oxygen species.71 (C) Iron chelation by siderophores can modulate host pathways, including the stabilization of HIF-1α, the upregulation of apoptosis genes and MAPK phosphatases, and the downregulation of DNA repair, mitosis, and cell cycle genes.61

3.1. Heavy metal sequestration and detoxification

Although they have extremely high affinity for ferric iron, siderophores are capable of binding additional metals as well.64,65 For example, DFO can bind Ga3+, Al3+ and In3+, albeit at a lower affinity than for Fe3+.66 Binding of non-iron metals by siderophores may benefit bacteria by preventing heavy metal toxicity.64 For example, pyoverdin and pyochelin-mediated binding of metals including Al3+, Co2+, Cu2+, Eu3+, Ni2+, Pb2+, Tb3+, and Zn2+ results in increased metal tolerance of Pseudomonas aeruginosa. This protection appears to be due to extracellular binding to metals by siderophores without subsequent internalization of the siderophore–metal complex.67–69 Although the pyochelin (Pch) receptor FptA binds Pch complexes with non-iron metals with similar affinity as Fe-Pch, these metals are internalized at substantially reduced rates or not at all.68 Ybt-mediated binding of Cu2+ results in increased copper resistance in Uropathogenic E. coli compared to rectal isolates, and Cu–Ybt complexes are present in urine from infected women at concentrations greater than Fe–Ybt.70 These findings indicate that siderophores can sequester toxic metals, and bacteria can exclude these complexes from their cytoplasm at the step of siderophore import. It is likely that this phenomenon is relevant to additional siderophores, heavy metals, and bacterial species.

3.2. Heavy-metal bound siderophores as protection against reactive oxygen species

In addition to preventing copper toxicity, Cu2+–Ybt complexes can act as a superoxide dismutase mimic, transforming oxygen radicals to more stable, less dangerous forms.71 In fact, Ybt from E. coli promotes intracellular survival in phagocytes that depends on both copper and the respiratory burst.71 The superoxide dismutase activity of Cu2+–Ybt complexes may facilitate the intracellular survival of some extracellular pathogens, including Uropathogenic E. coli, which have higher intracellular survival than non-pathogenic strains.71,72 Purified Ybt is also capable of inhibiting Reactive Oxygen Species (ROS) production in innate immune cells, particularly in neutrophils and macrophages.73 Aer and DFO were also shown to inhibit ROS production by neutrophils and macrophages, but at lower rates. The ability of apo-Ybt to inhibit ROS production was significantly greater than Fe–Ybt, suggesting that metal chelation is critical to this function. Whether the active ligand is Fe, Cu or another metal is unknown, and the mechanism of action of Ybt on phagocyte ROS production is unclear.73 These data indicate specific iron-independent abilities for Ybt that separate it from other bacterial siderophores and aid in virulence in strains of bacteria able to produce Ybt.

3.3. Uptake of non-iron metals

Similar to the requirement for ferric iron is the cellular requirement for nutritional Zn2+. The Y. pestis Zn2+ transporter ZnuABC is responsible for zinc uptake, however, znuBC mutants remain virulent in certain murine models of the plague.74 This has led to the suggestion of the presence of “zincophores,” molecules capable of scavenging zinc.75,76 A dedicated zincophore has yet to be discovered. However, two iron-binding siderophores have been shown to bind and utilize Zn2+ as well. Pseudomonas putida secretes pyridine-2,6-bis(thiocarboxylic acid) (PDTC), a small siderophore capable of binding and delivering Fe3+ and Zn2+ to the bacteria.77,78 In a Y. pestis znuBC mutant, Ybt is required for Zn2+ uptake and growth in Zn2+ limited conditions.79 This zinc transport requires the inner membrane receptor YbtX but not the ferric-Ybt receptor Psn. These data are the first to identify a role for dual metal-acquisition by a single siderophore and indicate a new role for siderophores during infection. It is likely that the binding of heavy metals other than iron by siderophores serves additional purposes for bacteria that are yet to be discovered.

3.4. Ability to disrupt host cell homeostasis

Since iron is imperative to many cellular processes, its chelation by siderophores could significantly affect host cellular homeostasis. Depending on the cellular response, siderophores can act as toxins causing cell death or immunomodulators. Such is the case with DFO, which has been evaluated as a cancer therapeutic because it can induce cell-cycle arrest and apoptosis, but can also induce inflammation via cytokine production in a variety of cell types.80–84 Microarray analysis of the respiratory epithelial cell response to Ent indicates profound changes in metabolism, marked by significant induction of MAPK phosphatase and apoptosis-associated genes, and repression of DNA replication, mitosis, DNA repair, and cell cycle-associated genes.61

Siderophores including Ent, Sal and Ybt induce the stabilization of the master transcription factor Hypoxia Inducible Factor-1α (HIF-1α).61,85 HIF-1α coordinates an adaptive response to hypoxia, which can occur directly from low oxygen or indirectly from low iron that compromises oxygen-dependent processes such as cellular respiration. Accordingly, HIF-1α stabilization is induced by either low oxygen or iron, as iron is a required cofactor for its degradation.86,87 It controls the transcription of inflammatory-cytokine, glycolysis, and angiogenesis genes. Stabilization of HIF-1α in the presence of Lcn2 is sufficient to induce secretion of the pro-inflammatory cytokine Interleukin-6 (IL-6), indicating that iron chelation by siderophores in the presence of Lcn2 has immunomodulatory properties.61 Accordingly, the combination of Ent and Lcn2 or Ybt and Lcn2 induces release of IL-6, IL-8 that is a powerful neutrophil chemoattractant,14,88 and CCL20 that is a lymphocyte chemoattractant.61 This inflammatory response was found to be due to cellular iron chelation by the siderophores, rather than the molecules themselves, and only occurred at high magnitude in the presence of Lcn2.61

It is unclear why such a specific set of cytokines would be upregulated by siderophore + Lcn2 stimulation. However, IL-6 may be an appropriate response since it is an acute phase reactant that can regulate iron homeostasis via hepcidin induction. In combination, IL-6 and CCL20 are important for development of the Th17 response and may initiate an adaptive response to infection with siderophore-producing bacteria such as K. pneumoniae. Lending validity to this hypothesis is research indicating the necessity of Th17 immunity in the host response to K. pneumoniae.89,90 Whether or not siderophore-mediated iron chelation modulates cytokine production in other cell types and the effect of siderophores bound to other metals, such as Cu–Ybt, on host cell homeostasis is unknown.

4. Implications of siderophore specialization during infection

As reviewed above, siderophores can be produced in a variety of structures, lending to variability in their affinity for iron, ability to be sequestered by Lcn2, ability to bind and utilize other heavy metals, and capacity to initiate host cellular pathways. Thus, the combination of siderophores produced by a bacterium could affect its pathogenesis, localization during infection, and instigation of specific immune responses (Fig. 3).

Siderophore production by bacteria can determine infection localization and “replicative niche”. (A) During infection with K. pneumoniae, the ability to produce Lcn2-evasive siderophores, such as Ybt, results in bronchopneumonia but not perivascular growth (left). In a Lcn2-deficient mouse, Ent promotes a perivascular localization of infection by accessing iron from transferrin (Tf) for growth.12,101 (B) Hypervirulent strains of K. pneumoniae that form pyogenic liver abscesses are associated with the secretion of Aer.100 (C) The receptors for Aer and Ybt are involved in bacteria colonization during Uropathogenic E. coli ascending UTI. In fact, vaccination with the Ybt receptor FyuA is protective for the kidney in murine models of ascending UTI.102,103

4.1. Ability of siderophores to evade Lcn2 mediates competition within the host

By using siderophores to acquire scarce host iron during infection, some bacteria can successfully compete for replicative niches in the intestines and the nasal cavity, allowing bacterial colonization. In fact, pathogens such as Salmonella enterica Serotype Typhimurium can exploit Lcn2 to outcompete other bacteria based solely on its ability to use Sal. Raffatellu et al. first demonstrated that S. Typhimurium can outcompete an isogenic mutant of the Sal receptor, IroN, in the intestines.91 Furthermore, by inducing IL-22 and expression of Lcn2, S. Typhimurium can outcompete commensal E. coli.92 Probiotics, such as E. coli Nissle, may protect against diarrheal infections. In this case, E. coli Nissle has a greater diversity of Lcn2-resistant siderophores than S. Typhimurium and causes a reduction in S. Typhimurium colonization that depends on the presence of Lcn2 and iron uptake by E. coli Nissle.93 In a nasal colonization model, Streptococcus pneumoniae and Haemophilus influenzae significantly induce Lcn2 but use siderophore-independent mechanisms to acquire iron.45 This inability to produce Ent and instead uptake iron through other siderophores could offer a competitive advantage by allowing colonization of a niche that is inhospitable to Enterobacteriaceae. By either not making siderophores or making more than a competitor, certain bacteria may leverage Lcn2 for their advantage.94

4.2. Ability to produce multiple siderophores correlates with clinical disease

K. pneumoniae is a Gram-negative pathogen that requires siderophores for virulence and is capable of causing pneumonia, urinary tract infections (UTI), wound infections, and septicemia.95 A comparison of isolates from various body sites of infected patients found increased prevalence of Ybt+ strains from respiratory tract samples when compared to urine, blood, and stool samples.12 Accordingly, Ybt production was sufficient to evade Lcn2 and cause pneumonia in a mouse model. Ent-dependent strains represent the predominate genotype of clinical isolates, but this genotype is inhibited in the wild-type mouse. In the Lcn2−/− mouse, this genotype is highly virulent. Therefore, decreased Lcn2 levels would be predicted to put patients at increased risk from infection. Indeed, lung Lcn2 levels vary significantly in intensive care patients, and low levels correlate with worsened survival from Gram-negative pneumonia.96 Similarly, Lcn2 levels rise in women with a urinary tract infection, and uropathogenic E. coli isolates with greater numbers of Lcn2-resistant siderophores are associated with greater bacterial density.47,48

A recent analysis of 57 carbapenem-resistant K. pneumoniae isolates of the epidemic sequence type (ST) 258 found an entire clade of ST258 strains with a deletion in the Ent exporter gene entS.97 The effect of this mutation is yet to be determined, but could represent another form of Lcn2 evasion.97 It is possible that these strains are “cheaters,” who utilize siderophores made by other bacteria for iron acquisition and energy usually reserved for Ent export for other functions, or acquire iron through siderophore-independent mechanisms. This could represent another situation in which bacteria utilize xenosiderophores to gain a competitive advantage in their replicative niche.

Whereas Ent may pose significant fitness tradeoffs, hypervirulent strains of K. pneumoniae causing severe liver abscesses in Asia are associated with carriage of multiple siderophore loci, including those for Ybt, Sal, and Aer.98 In particular, secretion of Aer is highly associated with pyogenic liver abscess (PLA) isolates of K. pneumoniae and results in significantly higher siderophore production compared to non-PLA strains.99,100 In contrast, only 2% of K. pneumoniae strains isolated from the respiratory tract, urine, blood or stool from a U.S. hospital carried Aer.12 It is possible that acquisition of Aer secretion could lead to more virulent strains of K. pneumoniae in the U.S. as well.

4.3. Siderophores can influence infection localization and determine bacterial “replicative niche”

The production of siderophores can influence the anatomic site and pattern of infection. In murine models of pneumonia, infection of wild-type mice with Ybt+ strains of K. pneumoniae resulted in bronchopneumonia with moderate bacterial density in the lungs and spleen.101 Infection with Ent+ strains in wild-type mice resulted in low bacterial burden and airway localization of inflammation. However, infection of Lcn2-deficient mice with an Ent+ or Ent+Sal+ mutant resulted in perivascular invasion by bacteria, high bacterial burden, invasion to the spleen, and poor survival.12,101 The perivascular space is rich with transferrin during infection, and the patterns of pneumonia correlated with the ability of Ent, but not Ybt, to support replication with transferrin as an iron source. This indicates that, due to differences in iron-scavenging capabilities between siderophores, Ent is required for perivascular infiltration but can only do so if Lcn2 is deficient. These data also suggest that Lcn2 could partially protect from bacterial dissemination.

The ability of E. coli to cause UTI is significantly affected by its complement of siderophores. Models of competitive infection in mice have shown that receptors for non-catecholate siderophores are instrumental in the establishment of bacterial colonies in ascending UTI infections.102 The receptors for Aer and Ybt, but not other siderophores, were shown to be critical for colonization of the bladder and the kidney.102 To further demonstrate the importance of these siderophore receptors, murine studies involving vaccination with the Ybt receptor resulted in protection from ascending UTI, specifically in the kidney.103 Successful vaccination with the Ybt receptor exemplifies the potential of siderophore-specific pathways as therapeutic targets during infection. Whether Aer or Ybt is more crucial in establishment of bladder or kidney infection has yet to be examined. However, Ybt genes are present among cystitis and pyelonephritis isolates at similar frequencies, suggesting that Ybt is equally important for infection in both the bladder and kidneys.104

Y. pestis requires the presence of Ybt for full virulence depending on the route of inoculation. Ybt mutants are fully virulent when infected intravenously in murine models of septicemic plague as long as the Znu zinc-transport system is intact.79 However, a Ybt mutant is essentially avirulent by subcutaneous injection, indicating the requirement of Ybt during a stage of bubonic plague infection.105 In a pneumonic plague model of infection, the absence of Ybt also decreased virulence of Y. pestis following intranasal infection with bacteria.106 The differential requirement for Ybt based on infection site may stem from a complex interaction with Fe, Cu and Zn. For example, whereas Zn can be acquired by redundant mechanisms in septicemic plague, perhaps Cu2+–Ybt or Fe3+–Ybt complexes are instrumental in pneumonic plague. These data exemplify the ability to siderophores to aid in virulence based on route of infection, and further support the hypothesis that siderophores can determine infection localization and replicative niche.

4.4. Siderophores, hypoxia, and disease

The ability of siderophores to induce host cell hypoxic responses can alter the outcome of infection. In a Caenorhabditis elegans infection model, P. aeruginosa killed worms in a pyoverdin-dependent manner that was associated with activation of a hypoxic host response. This response included activation of HIF-1α that was partially protective, as a hif-1 mutation exacerbated killing.107 Similarly, Y. enterocolitica induces HIF-1α during orogastric infection of mice, and HIF-1−/− mice have worsened survival.85 Furthermore, pharmacologic activation of HIF-1α can protect against infection through enhancement of phagocyte function.108 These results suggest that disruption of host cell homeostasis by siderophores can be lethal but can also serve as a danger signal that triggers a protective response to infection.

5. Conclusions

Significant progress has been made in the recent years to define the specialized characteristics of several siderophores, but many questions remain unaddressed. Siderophores are capable of binding non-iron heavy metals, but this has potential advantages and disadvantages to the bacteria. Recent work suggests that siderophores bind heavy metals to detoxify its environment. However, it is unclear how the bacterium avoids transporting the heavy metal–siderophore complex into the cell. Additionally, siderophores can act as both toxins and immunomodulators, but it is unclear whether activation of host pathways such as HIF-1α-regulated gene expression is ultimately beneficial to the host or the bacterium. Finally, it has been shown that strains producing certain siderophores have heightened virulence, as is the case with Aer-secreting strains of K. pneumoniae. Does the ability of Aer to preferentially chelate iron away from host cells and tissues directly cause its hypervirulence, or does Aer have other critical characteristics? These questions, among others, remain unanswered about the effects of siderophores during infection and colonization.

Although the first siderophores were discovered over half a century ago, the importance of siderophores as virulence factors during colonization and infection is still being uncovered today. Originally thought to be redundant iron-binding molecules, it is becoming more evident that individual siderophores can serve specific roles by evading Lcn2, binding non-iron heavy metals, initiating or inhibiting host cellular pathways, and determining the replicative niche during infection. Whether additional functions can be performed by distinct siderophores remains to be determined. Overall, more precise knowledge of the functions attributable to individual siderophores may allow for the development of novel therapeutics or vaccines, as is the case with the Ybt-receptor specific vaccine for murine ascending UTI.

References

Iron metabolism: from molecular mechanisms to clinical consequences

 ,  ,  ,  ,  ,  ,  ,  ,  ,  ,   and    

Proc. Natl. Acad. Sci. U. S. A.

  and    

Biochem. Biophys. Res. Commun.

  and    

Microbiol. Mol. Biol. Rev.

 ,  ,   and    

Proc. Natl. Acad. Sci. U. S. A.

 ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,   and    

 ,  ,  ,  ,  ,   and    

Proc. Natl. Acad. Sci. U. S. A.

 ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,   and    

Proc. Natl. Acad. Sci. U. S. A.

 ,  ,  ,  ,  ,   and    

Proc. Natl. Acad. Sci. U. S. A.

Indian J. Pathol. Microbiol.

Aqueous Solution Equilibrium and Kinetic Studies of Iron Siderophore and Model Siderophore Complexes

 ,  ,  ,  ,  ,  ,  ,   and    

Am. J. Respir. Cell Mol. Biol.

 ,  ,   and    

Front. Cell. Infect. Microbiol.

 ,  ,  ,  ,  ,  ,  ,  ,  ,  ,  ,   and    

Antimicrob. Agents Chemother.

Victoria Holden is a PhD candidate in the lab of Dr Michael A. Bachman at the University of Michigan. Her thesis work focuses on elucidating the host immune response to siderophores during infection with the bacteria Klebsiella pneumoniae. Victoria attained her Bachelors of Science degree in Immunology and Infectious Disease from the Pennsylvania State University in 2010. While there, she performed undergraduate research with Dr Margherita Cantorna. She is a member of the American Society for Microbiology and has served as President for the Organization of Microbiology and Immunology Students at the University of Michigan.

Michael Bachman is an Assistant Professor of Pathology and Clinical Director of Molecular Microbiology at the University of Michigan. His laboratory focuses on the mechanisms of pathogenesis in Klebsiella pneumoniae, particularly host–pathogen interactions centered on iron metabolism during pneumonia. He received his MD PhD from the University of Michigan Medical School in 2004, training in Microbiology and Immunology with Michele Swanson PhD. He was a resident in Clinical Pathology and fellow in Clinical Microbiology at the Hospital of the University of Pennsylvania, and a postdoctoral research fellow with Jeffrey Weiser MD.

© The Royal Society of Chemistry 2015

This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

What role do siderophores play in the survival of bacteria?

What is the value of siderophores?

How do siderophores function as virulence factors?

What is the meaning of siderophores?